host-identifier with IPv6

Ted Lemon Ted.Lemon at nominum.com
Wed Mar 4 21:00:02 UTC 2009


On Mar 4, 2009, at 1:44 PM, Frank Sweetser wrote:
> I'm basing that description of off the opinion David has already  
> expressed (an
> opinion which I happen to share).

No offense, but opinions are common as grass - what matters is not  
opinions, but evidence.   Show me the code, and I'll tell you whether  
or not it's a hack, and offer suggestions for improving it.   Or you  
can tell me why you think it's a hack, and I can suggest ways to  
mitigate it.   Or perhaps we will both agree, at that time, that it is  
in fact a hack, and there's nothing to be done about it other than a  
protocol enhancement.

> True, but then it's either a) present and functional, or b)  
> explicitly absent,
> as opposed to "should work, except when it doesn't".  Us  
> administrators like
> predictability =)

(b) is the status quo.   So you already have that predictability.    
Adding another variable makes things less predictable, not more  
predictable.




More information about the dhcp-users mailing list